"I do this thing where as soon as something is actually going well — a relationship, a job, anything — I find a way to ruin it. I'm not trying to. I just keep doing it."
If you've ever said this — to a friend, a therapist, or yourself in a quiet moment — you're not alone. The pattern of self-sabotage at moments of stability or success is one of the most consistently named felt experiences in adult life, and it's particularly painful because the conscious self genuinely wants the good thing while some other part of the system keeps undermining it.
The mechanism behind the pattern usually isn't about hating success or feeling unworthy. It's about anticipation tolerance — specifically, the relationship between how much discomfort the anticipation of loss produces and how much discomfort the realised loss would produce. For nervous systems that learned to expect loss, the anticipation can become more painful than the loss itself, and the sabotage often emerges as a way to convert the unbearable anticipation into a bearable known outcome. The system is choosing controlled loss over the dread of waiting for the loss to arrive on its own.
Key Takeaways
- Self-sabotage when things go well is usually about anticipation tolerance, not about hating success or feeling unworthy.
- The nervous system that learned to expect loss often finds anticipated loss more uncomfortable than realised loss.
- Sabotage in this context is often pre-emptive ending — choosing controlled loss over the dread of waiting.
- Anxious and disorganised attachment patterns make this dynamic more pronounced.
- Trauma history substantially shapes the pattern, particularly when good things were reliably followed by losses early in life.
- Pattern interruption usually requires building tolerance for good-thing-uncertainty rather than trying to convince yourself the good thing will last.
What's actually happening here?
Self-sabotage in moments of stability or success usually reflects an asymmetry between the nervous system's tolerance for anticipated loss and its tolerance for realised loss. For some people, sitting in the uncertainty of whether something good will continue is more painful than knowing for certain that it has ended. The sabotage emerges as a way to convert the unbearable uncertainty into a bearable known outcome — even when the bearable known outcome involves losing the good thing.
This isn't a conscious calculation. The system that produces the sabotage doesn't think "I'd rather end this now than wait for it to end on its own." The thinking, if it surfaces at all, is usually some version of "this isn't really that good anyway" or "I don't deserve this" or "this person isn't really the right person" — narratives that justify the sabotage rather than the actual mechanism that's driving it. The mechanism is more bodily than narrative: a nervous system that's calibrated for loss producing pre-emptive responses to good-thing-uncertainty.
The literature on this pattern, including work on what's sometimes called fear of good things or upper limit problems, has found consistent connection to early experiences where good things were reliably followed by losses, betrayals, or withdrawals. The nervous system that learned this association produces protective responses to good things in adulthood that the conscious self often experiences as inexplicable or self-defeating.
Mikulincer and Shaver's 2007 work on attachment in adulthood documents related patterns in attachment-influenced relationships, particularly the way that good relational experiences can trigger anxiety responses in people with insecure attachment patterns. The good thing isn't experienced as just good; it's experienced as good-but-also-the-setup-for-loss, and the protective responses to the second half of that experience can overwhelm the enjoyment of the first half.
Why doesn't it stop on its own?
The pattern persists because it functions, on the system's terms, even when it doesn't function on the conscious self's terms. The sabotage successfully resolves the unbearable anticipation by converting it into a known outcome. The system experiences this as relief, even when the conscious self experiences it as devastating loss. As long as the relief is registered, the pattern is being reinforced.
There's a related mechanism: the sabotage usually happens before the conscious self has time to recognise what's about to happen. The protective response operates on a faster timescale than the conscious recognition, which means that by the time the conscious self can name what's happening, the sabotage is often already in motion. The retrospective recognition of the pattern doesn't reach the system at the moment of action, which is why insight alone rarely changes the pattern.
The pattern also tends to be self-justifying. After sabotage, the conscious self often produces narratives that frame the lost good thing as not-really-that-good — the relationship had problems, the job wasn't right, the success wasn't what it appeared. The narratives reduce the cognitive dissonance of having sabotaged something good, but they also obscure the pattern by making each instance look like a justified individual decision rather than as an instance of a recurring pattern.
The cumulative effect can be substantial — years of good things sabotaged, opportunities ended early, relationships ended just as they were stabilising — without the person being able to see the full pattern, because each instance has its own narrative that makes it look like the right call at the time.
What pattern is underneath this?
The pattern under the pattern is usually some combination of attachment dynamics, learned associations from early life, and trauma-related implicit responses operating together. In our experience working with this question, the most common drivers fall into a few recognisable groups.
For people with anxious attachment patterns, good relational experiences often trigger anxiety about the eventual loss of the good experience, and the anxiety can become unbearable enough to drive sabotage of the relationship that produced it. The fuller picture of this dynamic is in anxious attachment.
For people with avoidant attachment patterns, intimacy that's actually working can trigger the protective distance the avoidant system relies on, producing sabotage that takes the form of pulling away precisely when closeness is going well. The fuller pattern is in avoidant attachment.
For people with disorganised attachment patterns, good things can produce particularly intense activation because the system simultaneously wants the good thing and fears it, and the resulting internal conflict can drive sabotage that looks erratic from outside but is actually the disorganised pattern producing predictable contradictions. The full picture is in disorganised attachment.
For people with childhood histories where good things were reliably followed by losses or betrayals (parental withdrawal of love after periods of warmth, sudden caregiving failures after periods of stability, abuse following affection), the nervous system often retains the implicit learning that good-thing-followed-by-bad-thing is the pattern to expect. Adult sabotage in this case is often the system trying to control the timing of the bad thing rather than wait for it to arrive on its own.
There are also non-trauma versions of the pattern. Some people sabotage because the good thing requires changes to identity that the system isn't ready for — a successful relationship requires becoming someone who has a successful relationship, a successful career requires becoming someone who has a successful career, and the identity shift can produce its own resistance. Some people sabotage because the good thing produces visibility they're not ready for. Some sabotage because they don't trust the good thing's source, even when the source is actually trustworthy. The patterns vary, and identifying which one is operating in your specific case matters for what intervention will help.
What's a tiny first move?
Pattern interruption usually starts with making the sabotage urge recognisable as an urge rather than as accurate information. The smallest useful first move is often noticing the moment the urge surfaces — the moment the mind starts producing reasons why the good thing isn't really that good, or the moment the body produces the impulse to do something that would undermine the good thing.
The recognition itself is the intervention. When you can name "this is the sabotage urge surfacing" rather than acting on the urge as if it were information, you create a small gap between urge and action that wasn't there before. The gap doesn't immediately stop the sabotage, but it creates space for a different response over time.
A useful second move is staying with the discomfort of the good-thing-uncertainty without resolving it. This is often hard, because the system has spent years learning to convert the discomfort into action. Sitting with the felt sense of "I have this thing and it might be lost and I don't know when or how" without doing anything about it is often the work that actually changes the pattern, and it usually takes substantial practice before it becomes possible.
A third move, particularly useful for trauma-influenced versions of the pattern, is identifying what the original learning was — what kind of good thing was followed by what kind of loss — and naming the connection between that early pattern and the current sabotage. The naming doesn't change the pattern by itself, but it often reduces the bewildered self-criticism that compounds the problem when the sabotage feels inexplicable.
The dynamic of how protective responses operate as defenses is explored further in boundaries vs walls, and the broader picture of how attachment patterns shape relationship dynamics is in how attachment theory helps relationships.
When this is bigger than self-help?
Some versions of this pattern are workable through personal work — recognition, gap-creation, slow building of tolerance for good-thing-uncertainty. Other versions involve trauma material that's difficult to reach without professional support. If the pattern is clearly tied to childhood experiences involving abuse, neglect, or sustained relational instability, working with a trauma-informed therapist is often substantially more effective than personal work alone, because the work that needs to happen is at the implicit-learning and somatic-response level rather than at the conscious-decision level.
If the sabotage involves substance use, self-harm, escalating destructiveness toward yourself or others, or persistent suicidal thinking, that's a clinical question and benefits from professional support beyond what self-help can provide.
The pattern isn't about hating good things. It's about a system that learned, somewhere, to expect their loss, and that finds the anticipation harder to bear than the loss itself. The work isn't to convince the system that good things won't end; it's to build the system's capacity to be in the uncertainty of good-things-existing without having to resolve it through pre-emptive ending. That capacity builds slowly, often with substantial discomfort along the way, and it usually changes the pattern more reliably than insight or willpower do.
Take the InnerPersona assessment — the assessment is designed to give you the words for what's been driving the sabotage pattern, including the attachment configuration and trait combinations most likely to be doing the work in your case.
Read next: Boundaries vs walls
Go deeper
Measure your own personality across 13 dimensions.
The InnerPersona assessment covers all 13 dimensions discussed in this article — free insights, no account required.
Frequently asked questions
Why do I sabotage things when they're finally going well?
Because for many people, anticipated loss is more uncomfortable than realised loss, and ending something on your own terms feels safer than waiting for it to be taken. When good things happen to a nervous system that learned to expect loss, the dread of the eventual loss can become more painful than just precipitating the loss now. The sabotage isn't really about not wanting the good thing; it's about not being able to tolerate the anticipation of losing it.
Is this an attachment thing?
Often yes, particularly for people with anxious or disorganised attachment patterns where good relational experiences activate fear about the eventual loss of those experiences. The activation can be intense enough that the relationship that's going well actually becomes harder to be in than a relationship with more familiar instability would be. Sabotage in this context is usually attempts to either restore the familiar instability or to end the relationship before the dreaded ending happens to you.
Is it about not feeling worthy of good things?
That's part of the picture for many people but not the whole story. Worthiness narratives often track with the sabotage pattern, but the underlying mechanism is usually more about anticipation tolerance than about actual self-evaluation. Even people who consciously feel worthy of good things can sabotage them when the nervous system response to good-thing-uncertainty is too uncomfortable to sit with. The worthiness reframe can help, but it doesn't always reach the somatic level the pattern operates at.
How do I stop self-sabotaging when good things happen?
The most useful work is usually building tolerance for the discomfort of good things being uncertain, rather than trying to convince yourself that good things won't end. The first part of this is often noticing the sabotage urge as a signal of nervous system activation rather than as accurate information about the relationship or situation. Naming the urge as the pattern rather than acting on it creates a small gap that wasn't there before, and over time the gap can widen.
Could this be a trauma response?
It often is, particularly when good things in childhood were reliably followed by losses, betrayals, or sudden withdrawals. The nervous system that learned to expect bad-thing-after-good-thing can produce the sabotage pattern automatically in adult life, even when the current situation has no actual signs of impending loss. Trauma-informed therapy is often particularly useful for this version of the pattern, because the work is at the implicit-learning level that conscious decision-making alone doesn't easily reach.
